/[pcre]/code/trunk/pcre_valid_utf8.c
ViewVC logotype

Diff of /code/trunk/pcre_valid_utf8.c

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 305 by ph10, Sun Jan 20 20:07:32 2008 UTC revision 1098 by chpe, Tue Oct 16 15:56:18 2012 UTC
# Line 6  Line 6 
6  and semantics are as close as possible to those of the Perl 5 language.  and semantics are as close as possible to those of the Perl 5 language.
7    
8                         Written by Philip Hazel                         Written by Philip Hazel
9             Copyright (c) 1997-2008 University of Cambridge             Copyright (c) 1997-2012 University of Cambridge
10    
11  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
12  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 54  strings. */ Line 54  strings. */
54  *************************************************/  *************************************************/
55    
56  /* This function is called (optionally) at the start of compile or match, to  /* This function is called (optionally) at the start of compile or match, to
57  validate that a supposed UTF-8 string is actually valid. The early check means  check that a supposed UTF-8 string is actually valid. The early check means
58  that subsequent code can assume it is dealing with a valid string. The check  that subsequent code can assume it is dealing with a valid string. The check
59  can be turned off for maximum performance, but the consequences of supplying  can be turned off for maximum performance, but the consequences of supplying an
60  an invalid string are then undefined.  invalid string are then undefined.
61    
62  Originally, this function checked according to RFC 2279, allowing for values in  Originally, this function checked according to RFC 2279, allowing for values in
63  the range 0 to 0x7fffffff, up to 6 bytes long, but ensuring that they were in  the range 0 to 0x7fffffff, up to 6 bytes long, but ensuring that they were in
64  the canonical format. Once somebody had pointed out RFC 3629 to me (it  the canonical format. Once somebody had pointed out RFC 3629 to me (it
65  obsoletes 2279), additional restrictions were applied. The values are now  obsoletes 2279), additional restrictions were applied. The values are now
66  limited to be between 0 and 0x0010ffff, no more than 4 bytes long, and the  limited to be between 0 and 0x0010ffff, no more than 4 bytes long, and the
67  subrange 0xd000 to 0xdfff is excluded.  subrange 0xd000 to 0xdfff is excluded. However, the format of 5-byte and 6-byte
68    characters is still checked.
69    
70    From release 8.13 more information about the details of the error are passed
71    back in the returned value:
72    
73    PCRE_UTF8_ERR0   No error
74    PCRE_UTF8_ERR1   Missing 1 byte at the end of the string
75    PCRE_UTF8_ERR2   Missing 2 bytes at the end of the string
76    PCRE_UTF8_ERR3   Missing 3 bytes at the end of the string
77    PCRE_UTF8_ERR4   Missing 4 bytes at the end of the string
78    PCRE_UTF8_ERR5   Missing 5 bytes at the end of the string
79    PCRE_UTF8_ERR6   2nd-byte's two top bits are not 0x80
80    PCRE_UTF8_ERR7   3rd-byte's two top bits are not 0x80
81    PCRE_UTF8_ERR8   4th-byte's two top bits are not 0x80
82    PCRE_UTF8_ERR9   5th-byte's two top bits are not 0x80
83    PCRE_UTF8_ERR10  6th-byte's two top bits are not 0x80
84    PCRE_UTF8_ERR11  5-byte character is not permitted by RFC 3629
85    PCRE_UTF8_ERR12  6-byte character is not permitted by RFC 3629
86    PCRE_UTF8_ERR13  4-byte character with value > 0x10ffff is not permitted
87    PCRE_UTF8_ERR14  3-byte character with value 0xd000-0xdfff is not permitted
88    PCRE_UTF8_ERR15  Overlong 2-byte sequence
89    PCRE_UTF8_ERR16  Overlong 3-byte sequence
90    PCRE_UTF8_ERR17  Overlong 4-byte sequence
91    PCRE_UTF8_ERR18  Overlong 5-byte sequence (won't ever occur)
92    PCRE_UTF8_ERR19  Overlong 6-byte sequence (won't ever occur)
93    PCRE_UTF8_ERR20  Isolated 0x80 byte (not within UTF-8 character)
94    PCRE_UTF8_ERR21  Byte with the illegal value 0xfe or 0xff
95    PCRE_UTF8_ERR22  Non-character
96    
97  Arguments:  Arguments:
98    string       points to the string    string       points to the string
99    length       length of string, or -1 if the string is zero-terminated    length       length of string, or -1 if the string is zero-terminated
100      errp         pointer to an error position offset variable
101    
102  Returns:       < 0    if the string is a valid UTF-8 string  Returns:       = 0    if the string is a valid UTF-8 string
103                 >= 0   otherwise; the value is the offset of the bad byte                 > 0    otherwise, setting the offset of the bad character
104  */  */
105    
106  int  int
107  _pcre_valid_utf8(const uschar *string, int length)  PRIV(valid_utf)(PCRE_PUCHAR string, int length, int *erroroffset)
108  {  {
109  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
110  register const uschar *p;  register PCRE_PUCHAR p;
111    
112  if (length < 0)  if (length < 0)
113    {    {
114    for (p = string; *p != 0; p++);    for (p = string; *p != 0; p++);
115    length = p - string;    length = (int)(p - string);
116    }    }
117    
118  for (p = string; length-- > 0; p++)  for (p = string; length-- > 0; p++)
119    {    {
120    register int ab;    register pcre_uchar ab, c, d;
121    register int c = *p;    pcre_uint32 v = 0;
122    if (c < 128) continue;  
123    if (c < 0xc0) return p - string;    c = *p;
124    ab = _pcre_utf8_table4[c & 0x3f];     /* Number of additional bytes */    if (c < 128) continue;                /* ASCII character */
125    if (length < ab || ab > 3) return p - string;  
126    length -= ab;    if (c < 0xc0)                         /* Isolated 10xx xxxx byte */
127        {
128        *erroroffset = (int)(p - string);
129        return PCRE_UTF8_ERR20;
130        }
131    
132      if (c >= 0xfe)                        /* Invalid 0xfe or 0xff bytes */
133        {
134        *erroroffset = (int)(p - string);
135        return PCRE_UTF8_ERR21;
136        }
137    
138      ab = PRIV(utf8_table4)[c & 0x3f];     /* Number of additional bytes */
139      if (length < ab)
140        {
141        *erroroffset = (int)(p - string);          /* Missing bytes */
142        return ab - length;                 /* Codes ERR1 to ERR5 */
143        }
144      length -= ab;                         /* Length remaining */
145    
146    /* Check top bits in the second byte */    /* Check top bits in the second byte */
   if ((*(++p) & 0xc0) != 0x80) return p - string;  
147    
148    /* Check for overlong sequences for each different length, and for the    if (((d = *(++p)) & 0xc0) != 0x80)
149    excluded range 0xd000 to 0xdfff.  */      {
150        *erroroffset = (int)(p - string) - 1;
151        return PCRE_UTF8_ERR6;
152        }
153    
154      /* For each length, check that the remaining bytes start with the 0x80 bit
155      set and not the 0x40 bit. Then check for an overlong sequence, and for the
156      excluded range 0xd800 to 0xdfff. */
157    
158    switch (ab)    switch (ab)
159      {      {
160      /* Check for xx00 000x (overlong sequence) */      /* 2-byte character. No further bytes to check for 0x80. Check first byte
161        for for xx00 000x (overlong sequence). */
162    
163      case 1:      case 1: if ((c & 0x3e) == 0)
164      if ((c & 0x3e) == 0) return p - string;        {
165      continue;   /* We know there aren't any more bytes to check */        *erroroffset = (int)(p - string) - 1;
166          return PCRE_UTF8_ERR15;
167          }
168        break;
169    
170      /* Check for 1110 0000, xx0x xxxx (overlong sequence) or      /* 3-byte character. Check third byte for 0x80. Then check first 2 bytes
171                   1110 1101, 1010 xxxx (0xd000 - 0xdfff) */        for 1110 0000, xx0x xxxx (overlong sequence) or
172              1110 1101, 1010 xxxx (0xd800 - 0xdfff) */
173    
174      case 2:      case 2:
175      if ((c == 0xe0 && (*p & 0x20) == 0) ||      if ((*(++p) & 0xc0) != 0x80)     /* Third byte */
176          (c == 0xed && *p >= 0xa0))        {
177        return p - string;        *erroroffset = (int)(p - string) - 2;
178          return PCRE_UTF8_ERR7;
179          }
180        if (c == 0xe0 && (d & 0x20) == 0)
181          {
182          *erroroffset = (int)(p - string) - 2;
183          return PCRE_UTF8_ERR16;
184          }
185        if (c == 0xed && d >= 0xa0)
186          {
187          *erroroffset = (int)(p - string) - 2;
188          return PCRE_UTF8_ERR14;
189          }
190        v = ((c & 0x0f) << 12) | ((d & 0x3f) << 6) | (*p & 0x3f);
191      break;      break;
192    
193      /* Check for 1111 0000, xx00 xxxx (overlong sequence) or      /* 4-byte character. Check 3rd and 4th bytes for 0x80. Then check first 2
194         greater than 0x0010ffff (f4 8f bf bf) */         bytes for for 1111 0000, xx00 xxxx (overlong sequence), then check for a
195           character greater than 0x0010ffff (f4 8f bf bf) */
196    
197      case 3:      case 3:
198      if ((c == 0xf0 && (*p & 0x30) == 0) ||      if ((*(++p) & 0xc0) != 0x80)     /* Third byte */
199          (c > 0xf4 ) ||        {
200          (c == 0xf4 && *p > 0x8f))        *erroroffset = (int)(p - string) - 2;
201        return p - string;        return PCRE_UTF8_ERR7;
202          }
203        if ((*(++p) & 0xc0) != 0x80)     /* Fourth byte */
204          {
205          *erroroffset = (int)(p - string) - 3;
206          return PCRE_UTF8_ERR8;
207          }
208        if (c == 0xf0 && (d & 0x30) == 0)
209          {
210          *erroroffset = (int)(p - string) - 3;
211          return PCRE_UTF8_ERR17;
212          }
213        if (c > 0xf4 || (c == 0xf4 && d > 0x8f))
214          {
215          *erroroffset = (int)(p - string) - 3;
216          return PCRE_UTF8_ERR13;
217          }
218        v = ((c & 0x07) << 18) | ((d & 0x3f) << 12) | ((p[-1] & 0x3f) << 6) | (*p & 0x3f);
219      break;      break;
220    
221  #if 0      /* 5-byte and 6-byte characters are not allowed by RFC 3629, and will be
222      /* These cases can no longer occur, as we restrict to a maximum of four      rejected by the length test below. However, we do the appropriate tests
223      bytes nowadays. Leave the code here in case we ever want to add an option      here so that overlong sequences get diagnosed, and also in case there is
224      for longer sequences. */      ever an option for handling these larger code points. */
225    
226        /* 5-byte character. Check 3rd, 4th, and 5th bytes for 0x80. Then check for
227        1111 1000, xx00 0xxx */
228    
     /* Check for 1111 1000, xx00 0xxx */  
229      case 4:      case 4:
230      if (c == 0xf8 && (*p & 0x38) == 0) return p - string;      if ((*(++p) & 0xc0) != 0x80)     /* Third byte */
231          {
232          *erroroffset = (int)(p - string) - 2;
233          return PCRE_UTF8_ERR7;
234          }
235        if ((*(++p) & 0xc0) != 0x80)     /* Fourth byte */
236          {
237          *erroroffset = (int)(p - string) - 3;
238          return PCRE_UTF8_ERR8;
239          }
240        if ((*(++p) & 0xc0) != 0x80)     /* Fifth byte */
241          {
242          *erroroffset = (int)(p - string) - 4;
243          return PCRE_UTF8_ERR9;
244          }
245        if (c == 0xf8 && (d & 0x38) == 0)
246          {
247          *erroroffset = (int)(p - string) - 4;
248          return PCRE_UTF8_ERR18;
249          }
250      break;      break;
251    
252      /* Check for leading 0xfe or 0xff, and then for 1111 1100, xx00 00xx */      /* 6-byte character. Check 3rd-6th bytes for 0x80. Then check for
253        1111 1100, xx00 00xx. */
254    
255      case 5:      case 5:
256      if (c == 0xfe || c == 0xff ||      if ((*(++p) & 0xc0) != 0x80)     /* Third byte */
257         (c == 0xfc && (*p & 0x3c) == 0)) return p - string;        {
258          *erroroffset = (int)(p - string) - 2;
259          return PCRE_UTF8_ERR7;
260          }
261        if ((*(++p) & 0xc0) != 0x80)     /* Fourth byte */
262          {
263          *erroroffset = (int)(p - string) - 3;
264          return PCRE_UTF8_ERR8;
265          }
266        if ((*(++p) & 0xc0) != 0x80)     /* Fifth byte */
267          {
268          *erroroffset = (int)(p - string) - 4;
269          return PCRE_UTF8_ERR9;
270          }
271        if ((*(++p) & 0xc0) != 0x80)     /* Sixth byte */
272          {
273          *erroroffset = (int)(p - string) - 5;
274          return PCRE_UTF8_ERR10;
275          }
276        if (c == 0xfc && (d & 0x3c) == 0)
277          {
278          *erroroffset = (int)(p - string) - 5;
279          return PCRE_UTF8_ERR19;
280          }
281      break;      break;
282  #endif      }
283    
284      /* Character is valid under RFC 2279, but 4-byte and 5-byte characters are
285      excluded by RFC 3629. The pointer p is currently at the last byte of the
286      character. */
287    
288      if (ab > 3)
289        {
290        *erroroffset = (int)(p - string) - ab;
291        return (ab == 4)? PCRE_UTF8_ERR11 : PCRE_UTF8_ERR12;
292      }      }
293    
294    /* Check for valid bytes after the 2nd, if any; all must start 10 */    /* Reject non-characters. The pointer p is currently at the last byte of the
295    while (--ab > 0)    character. */
296      if ((v & 0xfffeu) == 0xfffeu || (v >= 0xfdd0 && v <= 0xfdef))
297      {      {
298      if ((*(++p) & 0xc0) != 0x80) return p - string;      *erroroffset = (int)(p - string) - ab;
299        return PCRE_UTF8_ERR22;
300      }      }
301    }    }
302    
303    #else  /* Not SUPPORT_UTF */
304    (void)(string);  /* Keep picky compilers happy */
305    (void)(length);
306    (void)(erroroffset);
307  #endif  #endif
308    
309  return -1;  return PCRE_UTF8_ERR0;   /* This indicates success */
310  }  }
311    
312  /* End of pcre_valid_utf8.c */  /* End of pcre_valid_utf8.c */

Legend:
Removed from v.305  
changed lines
  Added in v.1098

  ViewVC Help
Powered by ViewVC 1.1.5